ABA Fundamentals

Delay discounting in the pigeon: In search of a magnitude effect

Holt et al. (2019) · Journal of the Experimental Analysis of Behavior 2019
★ The Verdict

Pigeons still refuse to wait even when they can see the bigger pile, so the missing magnitude effect is real and not just a comparison problem.

✓ Read this if BCBAs running delay-discounting or self-control protocols with non-verbal clients.
✗ Skip if Clinicians who only work with verbal adults or token economies.

01Research in Context

01

What this study did

Holt et al. (2019) tested whether pigeons would wait longer for bigger food rewards. They let birds choose between 8 pellets now or 32 pellets later in the same session.

The birds could see both amounts side-by-side before each choice. This setup removed any guess-work about reward size.

02

What they found

Even with the big and small piles right in front of them, the pigeons still picked the quick, tiny reward. Doubling the amount did not make them wait longer.

The birds acted as if 8 pellets and 32 pellets were worth the same once a delay was added.

03

How this fits with other research

Green et al. (2004) ran a similar test and also saw no magnitude effect, so the null result is not new. The 2019 study simply used a clearer within-session comparison.

Gowen et al. (2013) looks like a contradiction: those pigeons did show a magnitude effect. The difference is procedure. Emma used concurrent chains where delays grew inside the trial. Holt used adjusting amounts where delay was fixed and amount changed. The tasks measure different parts of choice.

Hilton et al. (2010) found the same null outcome in rats using the same adjusting-amount method, suggesting the procedure itself may hide magnitude effects in animals.

04

Why it matters

If you are teaching delay tolerance, do not assume bigger rewards will automatically buy more waiting. For humans we see magnitude effects; for pigeons (and maybe other non-verbal learners) the effect vanishes when the choice method makes them compare amount directly. Try mixing in other cues like signals or broken delays before you increase reinforcer size.

Free CEUs

Want CEUs on This Topic?

The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.

Join Free →
→ Action — try this Monday

Keep your delay-tolerance program, but add conditioned reinforcers or signals instead of simply offering more tokens.

02At a glance

Intervention
not applicable
Design
single case other
Finding
null

03Original abstract

The magnitude effect, where larger outcomes are discounted proportionally less than smaller outcomes, is a well-established phenomenon in delay discounting by human participants. To this point in the literature magnitude effects have not been reliably evidenced in nonhuman animals. , however, used a concurrent-chains arrangement with pigeon and found evidence for a magnitude effect. Grace et al. suggested that in many delay discounting experimental arrangements with nonhuman animals (e.g., adjusting amount, adjusting delay) the organism is not given the opportunity to directly compare outcomes of different sizes. They suggest that because of the lack of direct comparison it is difficult for the organism to determine the relative size of each outcome, which in turn mutes the effect of the amount differences between outcomes. As a test of this "comparison hypothesis," the present experiment was conducted to assess whether the magnitude effect would be evidenced in pigeon when using an adjusting amount procedure where outcomes of different amounts were presented proximally. In the present arrangement, pigeons were presented two choice panels in an operant chamber where each panel was associated with an independent adjusting amount delay discounting task, but with differing outcome amounts (i.e., a 32-food pellet panel and an 8-food pellet panel). In this arrangement the choice panels alternated in their availability within a session from trial block to trial block. The present findings indicate no reliable effect of amount, even when the outcomes were proximal and thus readily comparable. This result suggests that the lack of magnitude effect is not driven by the organism's ability to compare the difference in amount between choice alternatives.

Journal of the Experimental Analysis of Behavior, 2019 · doi:10.1002/jeab.515